Setting CLAUDE_CODE_RESTRICT_PERSONAL_CONFIG now stops your own skills, agents and hooks from widening permissions. In that mode, errors in your user settings file block every prompt and tool call until you fix them. You can set transcriptCacheTtl to 5m to ask for 5-minute prompt caching instead of the default 1 hour. claude mcp serve has a new --session-tunnel mode that connects through a relay. Idle conversations are now compacted shortly before their prompt cache runs out. Plugins can now send notifications through your terminal.
Several features are in this build but not switched on yet. Print mode can show each turn's text result as soon as that turn ends. MCP servers that failed with a retryable error can be retried at the start of the next turn. Images can be re-encoded as smaller WebP files when that saves space. Work on cloud sessions continues, with remote tool requests now pausing while the host computer sleeps. The fleet view can group and pin cloud sessions.
Tool search no longer freezes Claude Code when you have very many tools. A queued message that gets dropped is now handed back to you instead of being discarded. Prompts typed during a left-arrow hand-off to background agents are no longer lost. If a model rejects the 1M-token context option, Claude Code now leaves it out instead of failing every request. Plugin sync from claude.ai no longer removes plugin files that another running session is using. The end-of-turn message no longer includes token budget figures.
Written by our agent from the shipped bundle, not by Anthropic.
In cloud-remote sessions only, claude mcp serve can run over HTTP, dial a session tunnel, read tokens from a descriptor and return tool output objects
ExtensionsA host program can set CLAUDE_CODE_AUTOUPDATER_DISABLED_BY_HOST so plugin auto-updates stay off under the host's version pin
ExtensionsPlugin UI adds a ui.notify operation and Button text from children, and the Grep search now passes a searchOtherAppsData flag
Setting CLAUDE_CODE_RESTRICT_PERSONAL_CONFIG stops your own skills, agents and hooks from widening permissions, and makes key hooks required
Settings & config
CLAUDE_CODE_RESTRICT_PERSONAL_CONFIG mode restricts your own Claude Code filesSetting CLAUDE_CODE_RESTRICT_PERSONAL_CONFIG starts a session that restricts your own Claude Code files, alongside the other restricted modes
Want the reasoning? Read walks the 43 entries that probably matter to you, each one opening to what changed and why.
Read this release → Every row →19 more of these are in What probably matters to you, on page 1.
When settings do not reach a cloud session, Claude Code now names the broken settings file and tells you to fix it and reopen
Looking for one command inside a longer command line no longer ignores extra spaces, so commands that differ only in spacing stop matching
Unclear It is not established whether this check is used for permission rules.
Claude Code's check for sensitive file paths now lists stored credential files for git, the GitHub CLI and gcloud
Unclear It is not clear which mode the credential-file check applies in, or whether the list existed before this release.
Large _meta details in MCP tool results are now left out, with a higher limit for tools that show their own interface
Claude Code now strips every citation from a text block except web search result links when it processes Claude's replies
git stash list permission check no longer trusts arguments after --#The read-only rule for git stash list no longer assumes everything after -- on the command line is safe
Unclear It is not clear whether git stash list commands with arguments after -- now ask for your permission where they did not before.
The built-in web-fetch agent's guidance now says to read the page instead of answering from memory, and a follow-up hint is sometimes left out
The local MCP preview server now sends a nosniff header and applies its shared set of headers to more of its responses
A set of small changes touches auto mode messages, file search, image reading, teleport errors and usage tracking
Foreign artifact reads now always need consent, and reading comments always loads and checks the artifact first
Background tasks that want browser access now say to open the task on your computer, and a card notes that any answer skips the setup
MCP error redaction now limits its total work and stops with a message when the error repeats too much of the configuration
Pinning a cloud session now says "Can't pin a session that's running in another terminal"; cloud pinning and grouping are not switched on yet
Permission rules that contain no parentheses are now recognised as plain rules straight away
When cleaning terminal control codes out of output, Claude Code now checks each color or style code against a pattern before keeping it
Claude Code now builds its command list once and remembers results when checking messages for skill mentions
Unclear It is not stated under what condition the scan of your own skills folder now stops early.
Claude Code stops reading an MCP server's list if the server repeats a page marker, and no longer treats HTTP status errors as temporary
Syncing plugins from claude.ai now waits to remove old plugin folders that another running session is still using
When a request asking for the 1-million-token context window is refused, Claude Code can drop that request and send again
If a model rejects the 1M-token context option, Claude Code leaves it out for that model and tries again after /clear, compaction, resume or provider change
When a message you typed is queued and then dropped, Claude Code now hands the text back to you instead of discarding it
On Windows, plugin install records now match a project whose path is spelled differently or sits in another checkout of the same repository
Handing work to background agents with the left arrow now carries prompts you had typed, including ones taken back, into your input
In vim mode, ~ now stops at the end of the line, does nothing when there is nothing to toggle, and keeps the cursor in the right place
The Read tool now treats a blank page range as no page range, so the request no longer fails with a validation error
Disabling a plugin now looks across install locations when there is no exact match, and the artifact tool gains a new check
Unclear It is not stated what the artifact tool's new check rejects or when that mode applies.
When you switch to a different model mid-conversation, Claude Code now drops the prompt-cache record it was keeping for the old model
Claude Code now refuses to restore a prepared summary when that part of the conversation has already been compacted
Messages sent again after a reconnect are no longer dropped when they were waiting behind a permission prompt, and Stop also cancels waiting messages
The Agent SDK now passes many options as --flag=value, so a value starting with a dash is no longer mistaken for another flag
Background SessionStart hook output is skipped when Claude already has the same text from an earlier run, so the conversation is not padded with duplicates
Claude Code checks a second time before saying the working directory is gone, and tells the model when some Chrome tools are missing
If removing a message from a session file fails partway, Claude Code now rebuilds its in-memory index instead of ignoring the error
When a git command Claude Code runs internally fails, it is now logged and returned as a failed result instead of thrown as an error
Undoing a large paste now puts the text back at the cursor, and queued messages pulled back into the prompt box are marked as returned
Remote sessions now ignore connection errors and timeouts from a connection that has already been replaced by a newer one
Successful hook output is no longer dropped just because the hook ran somewhere else
Unclear It is not clear which hook event this applies to.
A switchable cleanup of trailing transcript messages no longer stops at empty SessionStart hook progress or success entries
Claude Code changed how it treats a plugin option whose value is not one of that option's allowed choices
Unclear It is not known whether invalid option values are now kept or dropped.
A check that decides whether the first queued command waits before starting no longer breaks when a session end has been requested
A tool result from an MCP server with empty content is no longer dropped if it carries metadata
A subagent is no longer moved to the background automatically while a tool call that must run alone is waiting behind it
After syncing skills, Claude Code now removes leftover hooks for skills that were removed, and logs a warning if that fails
When Claude Code falls back to another plugin entry, it now also considers entries whose project path matches where you are working
After a rewind, Claude Code now clears its stored last prompt so resuming does not bring back an old one
When Claude Code rewrites a block of text, citations that point at exact positions in it are now removed
A .git file with more than one hard link is no longer refused just for that when Claude Code checks it
Unclear It is not clear whether the new check still refuses a hard-linked .git file in some other way.
Hooks from synced skills are now cleaned up when sync turns off, the account changes or a skill goes away
Unclear When reads outside the working folder are blocked is not stated.
Plugin scope lookup and duplicate-scope filtering are fixed, and more git server errors are now recognised
If Claude Code fails to restart itself, the error is now logged and the terminal is reset instead of the error being printed raw
A dropped-request warning now also mentions over-long IDs, and the retry after re-enrolling a device now keeps the retry's result
Claude Code now tracks the column more accurately when expanding tabs in output, so alignment holds after color and style codes
3 more of these are in What probably matters to you, on page 1.
When a subagent edits a file, the Edit tool can attach a compact summary result, off unless switched on remotely
Unclear What the compact edit result changes in what you see when a subagent edits a file is not stated.
Cloud sessions now pause remote tool requests while the host computer sleeps and reconnect cleanly when it wakes
MCP servers that failed with a retryable error can be retried automatically at the start of each turn, for a limited number of turns
Claude Code can re-encode images as WebP when that is smaller, and now refuses images that are too costly to decode
Behind a switch that is off by default, auto mode registers tools reached through toolAliases or warns when an allow rule would skip its check
Unclear It is not clear how registering an aliased tool changes the way the classifier checks its calls.
Claude Code can now remind Claude about remote notifications that have waited longer than a set number of minutes
Claude Code can now count Claude's refusals since the conversation was last summarised, behind a setting that is off by default
Unclear What Claude Code does with this count when the setting is on is not stated.
Pressing Ctrl+C when no request is running no longer raises the stop-feedback notification, under a remotely controlled switch
Unclear It is not clear which setups the Cowork-related condition covers, so who sees this change is unknown.
A queued move to the background is now cancelled if you have text in the input box, and cleared when a turn stops
Claude Code keeps reading input during slow input hooks, and withdraws pending permission requests when you press Stop or the session ends
When staged files fail to land, Claude Code logs a warning and can tell the model the files may be missing or out of date
Claude Code has a new remote switch, read by the code that manages terminal sessions, with no visible feature behind it yet
Unclear What the tengu_cosmic_cookie switch controls is not known.
Claude Code now checks a new setting that the server can control, and what it does is not yet known
Unclear It is not known what this setting controls or where Claude Code uses it.
1 more of these is in What probably matters to you, on page 1.
home_dir becomes optional in the permission snapshot and is omitted in one mode, and remote-control joins the reserved directory names
The classifier that writes background-session status text adds a chatRelay mode with its own instructions and stops counting virtual messages
Unclear It is not clear what an active chat relay changes about the status shown.
Status messages for SDK users now carry a running count of summary characters written during automatic compaction
The labels Claude Code wraps around messages can now include who wrote the message, alongside reply and edit details
Under a certain condition, Claude Code now sets CLAUDE_CODE_RESTRICT_PERSONAL_CONFIG to 1 for the sessions it starts
Claude Code now probes the terminal for a new program-status feature (OSC 7501), treated as unsupported unless the terminal replies
Claude Code starts loading the client-data URL document during startup, and that document can now make a custom base URL count as Anthropic's own API
When the design-systems list request has its own stop signal, Claude Code now asks for an HTTP-only listing
Unclear It is not clear what HTTP-only changes about the lookup or what results it gives.
Remote thin-client sessions gain an onTurnStopped callback, and onTurnComplete now passes the message on when a turn ends
The text -p mode prints at the end of a turn and at exit, such as max-turns and budget messages, now goes through a shared printer
Unclear Whether the text printed in -p mode now looks different is not known.
The artifact five_class_asks capability is now a fixed value instead of depending on an environment variable or a remote flag
Unclear What this capability does for someone using Claude Code is not stated.
The propose_skills read check now logs when a skill was last shown and how many dropped lines never appeared in the conversation
The cleanup Claude Code runs after a sandboxed command was renamed and now takes a setting to skip its scrubbing step
Unclear What this changes for someone running commands in the sandbox is not clear.
Claude Code now waits on a second group of message attachments separately and does not hold the end of a turn for it
Unclear It is not stated which attached files belong to the second group.
On one proxied web search route, results for subagents now pass through an extra processing step before use
Unclear What the extra processing does to subagent search results is not known.
A remotely supplied override for claude-sonnet-4-6 is now skipped for some models that used to receive it
Unclear Which claude-sonnet-4-6 setups no longer get the override, and what the override changes, is not stated.
Claude Code now asks for the list of connected browsers by calling a tool, instead of reading a resource
Unclear Whether this changes anything you can see when using the browser integration is not known.
A Claude Code process that is shutting down no longer pre-warms or starts the background daemon, via a new standsDownAtExit option
Claude Code now keeps count of when it is in the middle of retrying after a temporary error
Claude Code now keeps track, per model, of models that refused the 1-million-token context window
The "Session can't start" error now carries a marker saying the session crashed
The prompt input now remembers when a draft has come back, keeps it across clears, and restores typed text handed back from the queue
Unclear It is not clear what you see differently when a draft is marked as having come back.
One event check now rejects events from the feedback survey, and a tool-limiting step now also reports which tools it kept
Claude Code now notes whether a dispatch was relayed or fell back, along with the cause and status code of a fallback
Usage reports from a home seed restore now note skipped answers, refused files, kept settings and left-out rules
Claude Code adds three new names to its usage-reporting stages, covering cloud home setup and shortened MCP results
In some sessions Claude Code now leaves out how long the process had been running when the session started
When a prompt filled in from a link is restored, Claude Code now notes that it came back, unless it was never submitted
Usage data now notes whether your git working copy had changes when the session started, and some usage data can be held back from sending
Claude Code now records when it compacts after a prompt-too-long error and when auto mode lets a rule allow a destructive MCP call
One way Claude Code looks up your username now returns nothing in some modes instead of reading GITHUB_ACTOR, USER or USERNAME
Unclear The condition under which the username is skipped is not known.
Usage data for the plugin dev-mods question now records each outcome through a lookup instead of only singling out "no one to ask"
Startup timing data now omits the process's absolute start time under one condition
Usage data sent when a plugin is fetched remotely now records which submodules were selected, recorded and fetched
Unclear Whether plugin installs now actually fetch git submodules is not confirmed.
Claude Code now counts a pull request as created when a curl command posts to a pull-request or merge-request address
Published verbatim by Anthropic for v2.1.295. Text is unmodified from the upstream changelog. Everything else on this page came out of the bundle instead, which is why the two lists don't match.
Of these 143 bullets, 35 name something an entry on this page also names, 32 name something no entry here does, and 76 name nothing specific enough to line up either way. The pairings are made on names both sides wrote down, a flag or a setting or a slash command, so read one as probably the same thing rather than as a fact, and read the middle number as candidates rather than as a miss count.
onFailure: "block" for command and HTTP hooks: a hook that can't start, times out, or exits with an unexpected code blocks the action instead of letting it through
No entry names this /copy picker, so a drafted message copies without its > markers
Probably copy-picker-now-lists-blockquotes-as-well-as-code-blocks claude plugin install, enable, disable and marketplace add when the settings file they write to does not load
Probably release-notes-bundle-for-21292, marketplace-add-warns-after-success claude -p run is waiting for when it stays open after its last turn
Probably print-mode-text-results-printed-at-each-turn-end timeouts.upstream_ttfb_ms on the Claude apps gateway's Bedrock, Vertex, Foundry and other cloud upstreams: a value you set now limits how long a stream may take to start there, after which it fails over or gets a 502
Probably gateway-upstream-ttfb-timeout-now-only-applies-when-operator, gateway-notes-upstream-ttfb-ms-now-applies-to-non-anthropic ← is waiting for the current tool to finish
No entry names this models list to every Claude apps gateway upstream: only the listed models are sent there, on failover too, and one * in an entry is a wildcard
Nothing to match on forceLoginMethod: "gateway" and forceLoginGatewayUrl in your own user settings on machines with no managed settings, so /login opens on that Claude apps gateway
No entry names this claude plugin validate when a plugin's README has no install line: it prints the line to paste and never changes the exit code, even with --strict
Probably claude-plugin-validate-adds-a-further-check-pass upstream_request_id to the Claude apps gateway's inference audit event: the request ID from Amazon Bedrock, the Anthropic API or another upstream, for support cases
Probably telemetry-additions-api-message-id-text-runs-joined-upstr $.ui.notify for mods: raises a native notification through your own notification setting and says which channel sent it
Probably plugin-uinotify-channel-and-tool-search-aware-description Button: strings and Text, so a row of a list is one pressable with a chip or a dim detail inside
No entry names this CLAUDE_CODE_RETRY_WATCHDOG_MAX_WAIT_MS to limit how long unattended retry mode (CLAUDE_CODE_RETRY_WATCHDOG) waits out 429 and 529 errors
Probably retry-watchdog-skip-can-be-overridden-by-claude-code-retry-w request-id header to the Claude apps gateway's successful inference responses, so request_id in Claude Code telemetry matches the gateway's audit log
Probably gateway-forwards-upstream-request-id, telemetry-additions-api-message-id-text-runs-joined-upstr [1m] model when a gateway, Bedrock, Vertex or Foundry refuses the context-1m beta; Claude Code now resends without it
No entry names this claude -p text output dropping earlier responses when background work started another turn; each turn's response now prints when the turn ends
Probably print-mode-text-results-printed-at-each-turn-end command_description instead of description
Probably bash-tool-input-command-description-is-accepted-as-descript host:80) to plain http:// pages on that host
No entry names this /plugin's Errors tab removing a marketplace that failed to load, and uninstalling its plugins, on Enter without asking first
Probably plugin-name-read-from-pluginjson-marketplace-clone-gains-p, plugin-errors-view-asks-for-confirmation-before-removing-mar, plugin-addressed-mcp-requestresponse-metadata-gated-behind claude plugin marketplace add reporting success for a marketplace whose name no plugin can be installed under; such an add is now refused
No entry names this CLAUDE_AUTO_BACKGROUND_TASKS moving a subagent to the background while an edit or shell command waited behind it, which started that call before the subagent finished
No entry names this claude remote-control sessions right after a phone message, without the /config toggle, or when started from inside Claude Code
Probably settings-panel-auto-update-channel-and-notification-changes, config-rows-commit-through-a-shared-writer-output-style-an #92661PushNotification reports "Remote Control inactive" (no_transport) in sessions served by `claude remote-control` Open
#99781PushNotification reports "Remote Control inactive" in sessions started with claude rc Closed
/tui exiting without a message, or with a raw system error, when Claude Code could not be started again
No entry names this availableModels not appearing in the /model picker for sessions signed in to a Claude apps gateway; such a model now shows its built-in row in place of a modelPicker row added for the same model
Probably remote-model-now-checks-whether-the-default-save-failed, model-pick-now-goes-through-a-saved-model-row-commit-with-s CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC hiding the Claude apps gateway spend limit in /usage and the status line. The spend limit request goes to the gateway the session is signed in to
Probably spend-limit-and-upsell-hint-text-refactored-into-a-lookup-ta, claude-mcp-serve-gains-an-http-transport-port-and-ses ←; Claude now keeps working in the foreground
Nothing to match on --tools and --restricted not applying to built-in tools that register after launch, and deprecated tool names reaching tools outside the caller's tool set
Probably sdk-spawns-cli-with-flagvalue-form-and-guards-values-st, tools-launch-list-now-withholds-unlisted-tools, new-claude-code-restrict-personal-config-env-var-recognized /loop that was moved to the background with ←; cancelling a pending wakeup now shows a notice
Probably resume-restores-a-pending-loop-wakeup-or-reports-it-as-mis FORCE_COLOR=3, which put color escape codes into the output Claude reads
No entry names this claude agents starting a new background service as it exited when both were stopped with a session open (for example at a reboot), which could delay shutdown and restart an interrupted session
Probably background-attach-prints-exit-notice-about-sessions-stopping #99369Agent view restarts the background daemon during system shutdown; the new daemon holds the shutdown for 90 s Open
worker being shown as "Agent" in auto-mode denial notices and in the activity lists of task detail views
No entry names this TaskStop, which can never be the cause
No entry names this /loop stopping without notice when its next wakeup came due while the session's process was down; the session now says so and Claude is told
Probably resume-restores-a-pending-loop-wakeup-or-reports-it-as-mis /advisor dialog showing a checkmark on a saved advisor model that is no longer available; it now opens on "No advisor"
No entry names this ~ moving the cursor past the last character of a line, so x after it did nothing, and 3~ running on into the next line
Nothing to match on .catch; such calls are now refused
Nothing to match on claude plugin test passing a session.append hook that removes tool call, tool result or thinking blocks that a real session keeps
No entry names this cat ran without printing it
Nothing to match on constructor or prototype always reading as their default and never reloading the plugin when edited
Nothing to match on /reload-plugins or session start was reading the mod's files
No entry names this /model, /fast and /output-style saving their setting without asking a plugin's config.set hook
Probably fast-mode-toggle-wrapped-for-remote-sessions-with-pre-switch, remote-model-now-checks-whether-the-default-save-failed, model-pick-now-goes-through-a-saved-model-row-commit-with-s claude mcp serve background Bash results not naming the output file, and its tool description promising a notification that never arrives
Probably mcp-tool-server-registers-extra-capabilities-when-serving-in, claude-mcp-serve-gains-an-http-transport-port-and-ses, mcp-serve-session-tunnel-and-tool-output-protocol #99934[BUG] `claude mcp serve`: Bash run_in_background output is unreachable since TaskOutput was removed (2.1.277) Open
CLAUDE_ENV_FILE not reaching the Bash tool after an in-app /resume or /branch
Probably restart-failure-messages #98933[BUG] CLAUDE_ENV_FILE env lost after in-app /resume — hook writes to startup session dir, Bash reads resumed dir (still on 2.1.287; #40391, #24775 auto-closed) Open
allowed-tools and effort being dropped when the Skill tool finished before the response stream ended, which denied the skill's Bash commands in -p runs
Probably new-claude-code-restrict-personal-config-mode-limits-what-a #99353[BUG] Skill allowed-tools rule is dropped when the Skill tool finishes before the response stream ends (2.1.289) Open
--plugin-dir plugin's folder in -p and SDK sessions; they are now written only where a mod is being developed
No entry names this /model saying a Max effort pick was saved as your default for new sessions; Max applies to the current session only
Probably remote-model-now-checks-whether-the-default-save-failed, model-pick-now-goes-through-a-saved-model-row-commit-with-s #99350[BUG] Clarify /model confirmation wording: max effort applies only to the current session Closed
pages parameter instead of treating it as omitted
Nothing to match on #98651[BUG] Read: `pages: ""` on a non-PDF file fails validation, and PreToolUse hooks can't work around it (still on 2.1.286) Closed
plugin install and uninstall at project or local scope on Windows missing the install record, or adding a second one, when only the drive letter's case differs
Probably release-notes-bundle-for-21292, plugin-install-lookup-refactored-into-a-shared-helper, plugin-install-records-matched-across-path-spellings-on-wind #92121Project-scope plugin installs are auto-created, never updated, and unreachable via --scope project (Windows drive-case duplicates) Open
↑ or Esc just after ←: prompt history now keeps it
Probably gateway-login-developer-sourced-url-and-esc-wording prompt.submit hook rewrote or dropped still being saved as typed to prompt history and to the transcript's queued-prompt records
No entry names this claude_code.auth OpenTelemetry login event not being emitted on Claude apps gateway sign-in. The event is exported when OpenTelemetry is configured on the machine or the session is already signed in to the gateway
No entry names this ~/.claude/seed-admin by an interrupted /ultrareview upload never being removed by the retention cleanup
No entry names this disableAutoMode was removed from settings
No entry names this /rewind being lost, and the removed turns coming back, when the session was moved to the background or resumed after being killed
Probably spend-limit-and-upsell-hint-text-refactored-into-a-lookup-ta session.receive hook asked for permission before passing a message on
No entry names this setMcpServers() as your own
No entry names this /config saving before a plugin's config.set hook was asked
Probably settings-panel-auto-update-channel-and-notification-changes, config-rows-commit-through-a-shared-writer-output-style-an /context and large file reads use: the gateway now gets them from AWS's CountTokens API instead of a one-token model request. Grant bedrock:CountTokens to use it
Probably bedrock-count-tokens-now-uses-aws-counttokens-with-backoff mcp__server__tool identifier
Probably auto-mode-resolves-toolaliases-for-mcp-tools-behind-tengu-so models: in Amazon Bedrock regions outside AWS's US geography. AWS always refused those requests; the gateway now tries the model in your own region, and a refusal names the model to add
Nothing to match on rate_limit_event usage-limit warnings to say whether the account has extra usage turned on
No entry names this plugin-authoring skill: it no longer tells a session with no terminal to run terminal commands, and explains sharing a mod only when asked
No entry names this -l, -c or -r flag now runs instead of failing the call
Nothing to match on claude plugin validate and plugin loading: when a hooks module is refused over a rebound top-level var, the error now names the line that rebinds it, the cause, and a fix
Probably claude-plugin-validate-adds-a-further-check-pass scriptPath refusal: it now says to pass the script inline via script, the route that works in sessions without a Read tool
No entry names this skills field, each once; a subagent with the Skill tool can still invoke the rest
Nothing to match on /loop wakeup alone, so pressing it twice detaches and the loop keeps running; press Esc to stop it
Probably resume-restores-a-pending-loop-wakeup-or-reports-it-as-mis claude agents stopped with its background service (macOS, or Linux without the service installed): running sessions now stop in about a minute unless it is run again, and a notice says so
Probably background-attach-prints-exit-notice-about-sessions-stopping MCP_PROTOCOL_NEGOTIATION=legacy opts out
No entry names this desktop policy key its bundled Claude Desktop schema doesn't know, so new Desktop settings need no gateway upgrade
Nothing to match on ws) MCP servers: a message over 16 MiB is no longer parsed and closes the connection, the limit the other transports already have
Nothing to match on CCR_AUTO_MODE_ALLOW, CCR_AUTO_MODE_ENVIRONMENT and CCR_AUTO_MODE_SOFT_DENY into a session's environment
Probably runner-spawns-child-with-ccr-auto-mode-env-vars-cleared CLAUDE_RUNNER_FETCH_SERVER_PROGRESS_CAP_MS tunes or turns off the wait
Probably git-fetch-gets-a-server-progress-stall-cap-env-var @Claude !restart had already confirmed the restart
No entry names this agent hook evaluations taking much longer at xhigh and max effort
Nothing to match on A model matched these bullets to the GitHub issues they fix, so a link can be wrong.
1 added and 1 removed, of 218 lines, about 22 words, in the prompt 14 of 27 arms receive. 4 other prompts also changed. The appended system-reminder blocks moved: 1 line added, 1 line removed.
Claude Code, interactive mode
13 prompt changes in this release could not be quoted from the build, so no entry on this page describes them.
603 documentation changes were recorded within 24 hours either side of this release, nearest first. The closest 12 are below. They're here because they happened near this release in time. That's not a claim that this release caused the edit, or that the page documents anything in it.
The 78 literal strings found in the bundle, with the number of entries that name each one. Picking one searches for it. A name is here because this build's code mentions it, which is not the same as it working or being finished.
What's wrong with this entry?